Article Dans Une Revue Computational Materials Science Année : 2024

Improved finite element algorithms for the thermo-metallurgical simulation of welding, additive manufacturing and related industrial processes

Résumé

Finite element simulations of welding, additive manufacturing and other thermomechanical processes require numerical algorithms both robust and economical in CPU time and memory resources. This paper focusses on thermal and metallurgical features of such simulations, leaving aside mechanical aspects. Two topics are envisaged: the evolution of the proportions of the various phases, resulting from anisothermal metallurgical transformations, and that of the temperature, depending on transformation latent heats. The simulation of phase changes uses Leblond and Devaux (1984)’s classical kinetic model for anisothermal transformations. With the hypothesis of modest variations of temperature over each discretized time-interval, the numerical algorithm proposed provides a quasi-exact solution within each such interval. The time-step may then be defined in reference to the sole characteristic time of the process itself, without considering those tied to the kinetic constants. The heat diffusion equation including latent heat effects is solved using a finite element technique suggested by Feulvarch et al. (2009), wherein the specific enthalpy is used as principal nodal unknown instead of the temperature. In this approach the temperature is considered as a well-defined function of the enthalpy (account being taken of the metallurgical transformations). This permits to use a robust and accurate algorithm for the integration in time, even in the presence of recalescence. Numerical examples are finally provided in order to illustrate the efficiency of the algorithms proposed.

Fichier principal
Vignette du fichier
Improved finite element algorithms-COMMAT-2024.pdf (4.41 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-04800575 , version 1 (24-11-2024)

Licence

Identifiants

Citer

Jean-Michel Bergheau, Jean-Baptiste Leblond, Yonggang Duan. Improved finite element algorithms for the thermo-metallurgical simulation of welding, additive manufacturing and related industrial processes. Computational Materials Science, 2024, 244, pp.113219. ⟨10.1016/j.commatsci.2024.113219⟩. ⟨hal-04800575⟩
46 Consultations
135 Téléchargements

Altmetric

Partager

  • More